Saharan Dust Composition on the Way to the Americas and Potential Impacts on Atmosphere and Biosphere

Saharan dust composition on the way to the Americas and potential impacts on atmosphere and biosphere

K. Kandler

Institut für Angewandte Geowissenschaften, Technische Universität Darmstadt, 64287 Darmstadt, Germany The Saharan dust plume 2 over the Western Atlantic Ocean

K. Kandler: Zum Einfluss des SaharastaubsKarayampudi et auf al., 1999, das doi : Klima10.1175/1520 – -Erkenntnise0477(1999)080 aus dem Saharan Mineral Dust Experiment (SAMUM)

The Saharan dust plume 3 on its arrival overover the the Caribbean Western Atlantic Ocean

daily averaged surface dust concentration (in µg-3), modelled for 23 June 1993

Kallos et al. 2006, doi: 10.1029/2005JD006207

K. Kandler: Zum Einfluss des SaharastaubsKarayampudi et auf al., 1999, das doi : Klima10.1175/1520 – -Erkenntnise0477(1999)080 aus dem Saharan Mineral Dust Experiment (SAMUM)

Case study of a dust event 4 from Bodélé depression reaching South America

daily lidar scans starting Feb 19, 2008 from Ben-Ami et al. 2010 doi: 10.5194/acp-10-7533-2010 5 Dust composition and its dependencies, interaction and potential impacts DEPOSITION TRANSPORT EMISSION

terrestrial solar radiative impact wet processing radiative impact photochemistry cloud impact indirect radiative impact anthropogeneous/ biomass burning aerosol admixture

heterogeneous chemistry

dry removal sea-salt interaction wet removal

terrestrial geological basement ecosystem impact type of weathering surface transport human/plant/animal marine ecosystem impact health issues chemical processing emission meteorology 6 Implications for radiation transfer – dust radiative forcing

Top-of the atmosphere (TOA) net 90% Clay 5% Quartz radiative forcing for different dust 5% Hematite mineralogical compositions(1) 75% Clays 20% Quartz 5% Hematite Shortwave forcing case study, including 80% Clays o measured aerosol size distribution o measured vertical aerosol distribution 20% Hematite o measured planetary surface albedo o measured particle shape cooling warming o derived particle refractive index -30 -20 -10 0 10

o derived particle single scattering albedo

(2) (2)

Planetary surface Top of the atmosphere

cooling

, W/m² , , W/m² ,

over ocean cooling

effects

warming

radiative radiative

Surface cooling over desert  decrease in sea surface temperature

 change in tropical cyclone activity(3)

aerosol aerosol

1data from Sokolik et al. 1999, doi: 10.1029/1998JD200048 2Otto et al. 2009, doi: 10.1111/j.1600-0889.2008.00389.x 3Evan et al. 2008, doi: 10.1029/2007GC001774 7 Implications for radiation transfer – dust composition and terrestrial radiation

Thermal infrared radiative forcing for different dust compositions(2)

² 40 m Mass extinction efficiency for different minerals /

W 30

in the atmospheric window wavelength region(1) ,

g n

i 20

r o

f 10

A O

T 0

² m

/ 10

dust

storm

(Niger) x

dust g

30 (Barbados)

dust

Afghan Afghan

SW USA SW USA l

e 40

dust

n Negev Negev

w 50

Saharan d

e 60

Saharan

f r

u 70 S 80

1Hansell et al. 2011, doi:10.5194/acp-11-1527-2011 2data from Sokolik et al. 1998, doi: 10.1029/98JD00049 8

Dust composition and clouds

Modification by clouds . Uptake of gaseous precursor species via aqueous chemistry . Aqueous reaction with (acidic) species . Composition-selective removal by wet deposition . Internal mixing by droplet/particle scavenging or coalescence

Impact on clouds . Fresh dust particles can act as cloud condensation nuclei (dependency on mineralogy)(1) . Dust particles as giant cloud condensation nuclei (GCCN) may alter precipitation(2) . Increase of ice nucleus (IN) concentrations, IN at higher temperatures(3) . Mineral dust contributes by 21 to 84% to the IN concentration(4) over the Amazon basin . In a case study, more than 79 % of cloud droplets at Cape Verde contained dust particles(5) . Water vapor competition of GCCN versus smaller dust particles makes precipitation impact depending on cloud conditions (i. e. liquid water content, but also gaseous chemistry)(6)  impact is ambiguous, depending on cloud microstructure(7) . Short-lived convective clouds are most sensitive(6) . Indirect impact through change of atmospheric dynamics by radiative impact (surface cooling, increase in atmospheric stability)(6)

1Kumar et al. 2011, 10.5194/acp-11-3527-2011 2Yin et al. 2000, 10.1016/S0169-8095(99)00046-0 3Sassen et al. 2003, doi: 10.1029/2003GL017371 4Prenni et al. 2009, doi: 10.1038/NGEO517 5Twohy et al. 2009, doi: 10.1029/2008GL035846 6Rosenfeld et al. 2001, doi: 10.1073/pnas.101122798 7Seifert et al. 2011, doi: 10.5194/acpd-11-20203-2011 9

Dust and ecosystems

Marine ecosystems . Fe (and possibly, P) can limit bio-productivity on Oceans directly or by co-limiting N fixation(1,2) . Saharan dust is made responsible for the degradation of coral reefs(3), but possible pathways are still explored(4) . Toxic red tides in the Gulf of Mexico need dust Fe (and maybe P) input to start(5) . On the nature (and impact) of P on marine ecosystems “remarkably little is known“(2) . P input by dust can increase bacterial activity in Mediterranean freshwater ecosystems(6) Terrestrial ecosystems . Forest ecosystems on extremely leached soils (like Amazonia) are short in nutrients (P, K) which can be provided by dust fall(7,8) . For example, half of the total inputs to soil-and-biomass P can be derived from dust in Puerto Rico’s Luquillo Mountains(9) . Forests on less leached soils can have deficit of Ca and K(10) . A tropical Andean forest in Ecuador receives considerable amounts of Ca and Mg from Saharan dust(11)

1Jickells et al. 2005, doi: 10.1126/science.1105959 2Okin et al. 2011, doi: 10.1029/2010GB003858 3Shinn et al. 2000, doi: 10.1029/2000GL011599 4Rypien 2008, doi: 10.3354/meps07600 5Walsh et al. 2006, doi: 10.1029/2004JC002813 6Reche et al. 2009, doi: 10.4319/lo.2009.54.3.0869 7Swap et al. 1992, doi: 10.1034/j.1600-0889.1992.t01-1-00005.x 8Okin et al. 2004, doi: 10.1029/2003GB002145 9Pett-Ridge 2009, doi: 10.1007/s10533-009-9308-x 10Bond 2010, doi: 10.1007/s11104-010-0440-0 11Boy et al. 2008, doi: 10.1029/2007GB002960 10 Emission stage – sources, composition, variation DEPOSITION TRANSPORT EMISSION

terrestrial solar radiative impact wet processing radiative impact photochemistry cloud impact indirect radiative impact anthropogeneous/ biomass burning aerosol admixture

heterogeneous chemistry

dry removal sea-salt interaction wet removal

terrestrial geological basement ecosystem impact type of weathering surface transport human/plant/animal marine ecosystem impact health issues chemical processing emission meteorology 11

Emission stage – general mineralogy

. Dust composition is highly variably . Quartz and phyllosilicates are omnipresent . Phyllosilicates might be . (frequently reported) kaolinite, illite . (less frequently) chlorite, muscovite, montmorillionite, biotite, palygorskite, smectites and inter-stratified clay minerals . Mostly, additional silicates are reported . (frequently) albite, anorthite, K-feldspars . (less frequently) chrysotile, orthoclase . Calcite, dolomite and sometimes apatite are found in varying abundance . Hematite, goethite and sometimes ilmenite are the main iron compounds . Sulfates, nitrates and chlorides are usually not reported with their mineralogical denomination (some of them might [fractionally] recrystallize depending on environmental conditions) . In addition, a plethora of other mineral species are reported, including biological debris (diatomite), metal oxides (rutile, periclase, baddeleyite, spinel), iron-rich minerals (lepidocrocite, limonite), carbonates (aragonite, magnesite), sulfates (anhydrite, gypsum, thenardite, mirabilite, mascagnite, glauberite), silicates (chloritoid, leucite, forsterite, zircon, enstatite) and graphite 12

Emission stage – sources and composition

1.78-2.41Bu Composition of 0.74-2.97To1 Saharan dust and 2.11BL (SD) topsoils characterized by (Ca+Mg)/Fe ratio(1) 0.40-0.46Kr (SD) 2.30-2.53Av1 5.17GF Li Av2 3.77 0.62 2.08Av2 6.07Li (SD) 3.65Li Ca 3.56 1.16Av1 3.35-7.24Mo 4.29-8.40Ab GT 0.65Av2 0.71-1.19 CD 2.56 0.66Mo 0.61-0.84Av1 0.66Ca 0.88-2.12To2 0.62-0.85GT

Br He 0.25-2.17 Mou 0.22 Mo 0.50 Li 0.45 Mou 0.53 Mo 0.63 0.37 Mou Sh Mo Ca 0.61 0.86-1.39 0.43 0.45 Mou 0.42Ca 0.74 0.40-0.65Or Ra Major source 0.83 0.23Mb 0.80-0.83Wi (Ca+Mg)/Fe regions compiled 0.42MW Wi dust 0.66-0.84 0.25 0.5 from different <0.25 1 - 2 2 - 4 >4 techniques are 0.04-0.21Ad - 0.5 - 1 identified(2) soil

1Scheuvens et al., in preparation for Earth. Sci. Reviews – details on original literature are given there 2Formenti et al. 2011, doi: 10.5194/acp-11-8231-2011

13 Mineralogical composition as function of source regions – an example

 dust intensity  Quartz K-feldspar Plagioclase Calcite Illite (1M) I Kaolinite Smectites Chlorite Gypsum Halite Hematite K

Rutile

Mai13 Mai14 Mai15 Mai16 Mai17 Mai18 Mai19 Mai20 Mai21 Mai22 Mai23 Mai24 Mai25 Mai26 Mai27 Mai28 Mai29 Mai30 Mai31

Jun01 Jun02 Jun03 Jun04 Jun05 Jun06 dust phases by meteorology and satellite obs.  DU1 DU2 DU3 Quartz K-feldspar Plagioclase Calcite Illite (1M) dominant Kaolinite major Smectites minor Chlorite traces Gypsum detected Halite not det. Hematite not invest.

Rutile

Jan14 Jan15 Jan16 Jan17 Jan18 Jan19 Jan20 Jan21 Jan22 Jan23 Jan24 Jan25 Jan26 Jan27 Jan28 Jan29 Jan30 Jan31

Feb 01Feb 02Feb 03Feb 04Feb 05Feb 06Feb 07Feb 08Feb 09Feb

data from Kandler et al. 2009, doi: 10.1111/j.1600-0889.2008.00385.x and Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 14

Change in mineralogical composition with particle size seen by chemical fingerprints

0.5 to < 1 µm 1 to < 2.5 µm 2.5 to < 10 µm 10 to < 25 µm 0.6 0.6 0.6 0.6

0.5 0.5 0.5 0.5

i 0.4 0.4 0.4 0.4

) e

F 0.3 0.3 0.3 0.3

+ g

M 0.2 0.2 0.2 0.2 (

0.1 0.1 0.1 0.1

0.0 0.0 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.6 0.6 0.6 0.6

0.5 0.5 0.5 0.5

S /

) 0.4 0.4 0.4 0.4

a C

+ 0.3 0.3 0.3 0.3

K +

a 0.2 0.2 0.2 0.2

N ( 0.1 0.1 0.1 0.1

0.0 0.0 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Al/Si Al/Si Al/Si Al/Si 0.03 0.01 0.003 0.001 0.0003 0.0001

data from Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 15

Change in mineralogical composition with particle size seen by chemical fingerprints

0.5 to < 1 µm 1 to < 2.5 µm 2.5 to < 10 µm 10 to < 25 µm 0.6 0.6 0.6 0.6

0.5 0.5 0.5 0.5

i 0.4 0.4 0.4 0.4

) e

F 0.3 0.3 0.3 0.3

+ g

M 0.2 0.2 0.2 0.2 S ( S S S 0.1 0.1 0.1 0.1

0.0Q C 0.0Q C 0.0Q F 0.0Q F 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.6 0.6 0.6 0.6

0.5 0.5 0.5 0.5

S /

) 0.4 0.4 0.4 0.4

a C

+ 0.3 0.3 0.3 F 0.3 F

K +

a 0.2 S 0.2 0.2 0.2 N ( S 0.1 0.1 0.1 S 0.1 S C 0.0Q C 0.0Q 0.0Q 0.0Q 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Al/Si Al/Si Al/Si Al/Si 0.03 0.01 0.003 0.001 0.0003 0.0001

data from Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 16

Change in mineralogical composition with particle size seen by chemical fingerprints

0.5 to < 1 µm 1 to < 2.5 µm 2.5 to < 10 µm 10 to < 25 µm 0.6SE on C-K 0.6 0.6 0.6 0.5 0.5 0.5 0.5

i 0.4 0.4 0.4 0.4

) e

F 0.3 0.3 Q3 0.3 0.3

+ Ti g

M 0.2 0.2 0.2 0.2 S ( S Q2 S S 0.1 0.1 Si 0.1 0.1 0.0Q C 0.0Q P C 0.0Q F 0.0Q F 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.6 Fe 0.6 0.6 0.6

0.5 0.5Si 0.5 0.5

i Mg

S /

) 0.4 0.4 0.4 0.4 a

C Q1

+ 0.3 0.3 0.3 F 0.3 F

K +

a 0.2 S 0.2 0.2 0.2 N ( S 0.1 0.1 0.1 S 0.1 S S C 0.0Q C 0.0Q 0.0Q 0.0Q 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Al/Si Al/Si Al/Si Al/Si 0.03 0.01 0.003 0.001 0.0003 0.0001

data from Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 17 Bodélé depression – “a single spot”(1)?

10 0 km

1Koren et al. 2006, doi: 10.1088/1748-9326/1/1/014005 Image from Chappell et al. 2008, doi: 10.1029/2007JD009032 18 Bodélé depression – “a single spot”(1)?

Data from Bristow et al. 2010, doi: 10.1029/2010GL043486

km 2 10 0 km

Fe/Ca ratio

1 2 3 4

1Koren et al. 2006, doi: 10.1088/1748-9326/1/1/014005 Image from Chappell et al. 2008, doi: 10.1029/2007JD009032 19 Variability across the Ocean – Fe/Ca as a tracer

Atomic ratio of Fe/Ca and its geometric standard deviation

0.72 2.10 1.22 (1) 0.95 0.53 1.77 1.19 1.71 (6) (4) (9)

1.05 (8) 1.31 1.48 2.13 (2) 1.06 / 1.93 1.60 advected local (7) 1.30 / 1.13 1.40 (5) 1.72 2.18 (3) (10)

1Kandler et al. 2009, doi: 10.1111/j.1600-0889.2008.00385.x 2Bristow et al. 2010, doi: 10.1029/2010GL043486 3Klaver et al. 2011, doi: 10.1002/qj.889 4DUST, Paris et al. 2010, 10.5194/acp-10-4273-2010 5Formenti et al. 2008, doi: 10.1029/2008JD009903 6Kandler et al. 2007, doi: 10.1016/j.atmosenv.2007.06.047 7Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 8Formenti et al. 2003, doi: 10.1029/2002JD002648 9Reid et al. 2003, doi: 10.1029/2002JD002935 10Formenti et al. 2001, doi: 10.1029/2000JD900827 20 Variability across the Ocean – Fe/Ca as a tracer

v 2.0 e

d Fe/Ca

t s

v 1.0

e o

Atomic ratio of Fe/Ca andd its geometric standard deviation

Fe/Ca e 0.8 g

d 2.0 t

+ 0.6

o a

e 0.4 e

g 1.0 m

+ 0.8 0.72

n e

a 0.6 g e 2.100.2

m 1.22

0.4 (1) 1 0.9510 100 o

0.53 e g 1.77 1.19"PM"- x 1 10 100 1.71 (6) (4) (9) "PM"-x

1.05 (8)

v 1.31 e

d 4.0 Fe/Ca d

t 1.48 s (2)

o 2.13

e 2.0

g 1.06 / 1.93

+ 1.60 advected local

n (7) a 1.0 e 1.30 / 1.13 1.40

m 0.8 (5)

o 0.6 1.72 e

g (3) 2.18 1 10 100 (10) "PM"-x

1Kandler et al. 2009, doi: 10.1111/j.1600-0889.2008.00385.x 2Bristow et al. 2010, doi: 10.1029/2010GL043486 3Klaver et al. 2011, doi: 10.1002/qj.889 4DUST, Paris et al. 2010, 10.5194/acp-10-4273-2010 5Formenti et al. 2008, doi: 10.1029/2008JD009903 6Kandler et al. 2007, doi: 10.1016/j.atmosenv.2007.06.047 7Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 8Formenti et al. 2003, doi: 10.1029/2002JD002648 9Reid et al. 2003, doi: 10.1029/2002JD002935 10Formenti et al. 2001, doi: 10.1029/2000JD900827 21 Transport stage – selective removal and admixture DEPOSITION TRANSPORT EMISSION

terrestrial solar radiative impact wet processing radiative impact photochemistry cloud impact indirect radiative impact anthropogeneous/ biomass burning aerosol admixture

heterogeneous chemistry

dry removal sea-salt interaction wet removal

terrestrial geological basement ecosystem impact type of weathering surface transport human/plant/animal marine ecosystem impact health issues chemical processing emission meteorology 22 Transport stage – selective removal and admixture

By sedimentation . The largest particles > 50 µm are quickly removed  relative abundance of quartz and feldspars decreases, that of clay minerals increases (carbonate contents are usually not strongly impacted) . Direct result: soil composition does not necessarily reflect generated dust composition, even close to source

By admixing . For example, sulfate and soot particles may be added do dust aerosol (or vice versa) and form an external mixture, which than can affect its radiative properties

By selective wet deposition . Dust particles containing larger amounts of soluble material may be preferentially removed by rain-out / washout . Dust particles more sensitive to chemical processing (i. e. carbonates to nitric acid) may quickly grow into large droplets under humid conditions and can be removed

Average 1.0 23 e

c 0.8

Chemical n

a d

composition n 0.6 u

b other at different a

e 0.4 carbonaceous

v i

locations t sulfates a l 0.2

e silicate sulfate mixtures r Morocco(1) (source-specific) silicates 0.0 quartz 0.1 1 10 100 sodium chloride 1.0

gypsum e

c 0.8 other calcium-rich n

a carbonates d

n 0.6 titanium-rich u

b iron-rich Tenerife(2) a depletion

e 0.4

v i

(SAL transport) t a

l 0.2

e r 0.0 0.1 1 10 100 mixing1.0

other e

c 0.8 soot mixtures n

a admixture sulfates d

(3) n 0.6 silicate sulfate mixtures Cape Verde u b silicates a

(MBL transport) e 0.4 quartz

v i

t sodium chloride a l 0.2

e carbonates r oxides 0.0 1Kandler et al. 2009, doi: 10.1111/j.1600-0889.2008.00385.x 2Kandler et al. 2007, doi: 10.1016/j.atmosenv.2007.06.047 3Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 24

Transport stage – processing and mixing

DEPOSITION TRANSPORT EMISSION

terrestrial solar radiative impact wet processing radiative impact photochemistry cloud impact indirect radiative impact anthropogeneous/ biomass burning aerosol admixture

heterogeneous chemistry

dry removal sea-salt interaction wet removal

terrestrial geological basement ecosystem impact type of weathering surface transport human/plant/animal marine ecosystem impact health issues chemical processing emission meteorology 25

Transport stage – processing

Processes affecting dust composition . Uptake of non-dust species by mechanical mixing (sea-salt mixture not found at Cape Verde(1,2), but reported in America(3); mixing seems most efficient for the transition size range 0.5 to 2 µm particle diameter) . Condensation of secondary species (e. g., organics) on the dust surface(4) . Reaction of dust with (acidic) species, depending on dust composition (e. g., calcic vs. silicic)(5) (6,7) (8,9) . SO2 can be oxidized in dry state on dust surface in presence of ozone , but effect saturates

. Many pathways in aqueous state from SO2 to sulfate, efficiency depending on conditions (like gaseous concentrations, cloud water pH, photochemistry...)(5,10)

Resulting changes in dust properties . Increase in sedimentation removal by increase in particle size(11) . Deposition of hygroscopic matter on dust particles increases CCN ability(12) . Sulfuric acid/sulfate and some organic matter decrease IN ability, but strength of effect depends on mineralogy(13,14) . Changes in solubility of nutrients

1Dall‘Osto et al. 2010, doi: 10.1016/j.atmosenv.2010.05.030 2Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 3Worobiec et al. 2007, doi: 10.1016/j.atmosenv.2007.07.056 4Deboudt et al. 2010, doi: 10.1029/2010JD013921 5Cwiertny et al. 2008, doi: 10.1146/annurev.physchem.59.032607.093630 6Usher et al. 2002, doi: 10.1029/2002JD002051 7Ullerstam et al. 2002, 10.1039/b203529b 8Goodman et al 2001, doi: 10.1021/jp004423z 9Manktelow et al. 2010, doi: 10.5194/acp-10-365-2010 10Seinfeld & Pandis, Wiley 2006 11Zhang 2008, doi: 10.1111/j.1600-0889.2008.00358.x 12Gibson et al. 2007, doi: 10.1080/02786820701557222 13Cziczo et al. 2009, doi: 10.1088/1748-9326/4/4/044013 14Eastwood et al. 2009, doi: 10.1029/2008GL035997 26 Dependence of the chemical reactivity on particle material over Niger

1.0

sulfate nitrate chlorine

e i

t 0.8 h

c clear-sky

w e

n e i

d 0.6 in-cloud

d r

f 0.4

r u

p m

m 0.2

n c

0.0

l l l l l l l l l

) ) )

i i i

z z z

i i i

o a t o a t o a t

S S S

T T T

D C D C D C

Q Q Q

e e e

F F F

( ( (

l l l

i i i

S S S

data from Matsuki et al. (2010), 10.5194/acp-10-1057-2010 27 Dependence of the chemical reactivity on particle material over Niger

1.0

sulfate nitrate chlorine

e i

t 0.8

h large differences

c clear-sky e w in mixture

t  mixing by

n e i reaction or in d 0.6 in-cloud

n aqueous phase,

s o

a not mechanically

d r

f 0.4

r u e

reactivityo least dependant on b

particlep material, maybe due increase of mixing by m

Ca sulfatem 0. being2 less u o cloud processing

hygroscopic than Ca nitrate

n c and chloride, constraining evident for hygroscopic in-depth processing, and/or species, but also for

due to dry0 .sulfate0 formation silicate particles

l l l l l l l l l

) ) )

i i i

z z z

i i i

o a t o a t o a t

S S S

T T T

D C D C D C

Q Q Q

e e e

F F F

( ( (

l l l

i i

i in general, also for clear-sky conditions reactivity

S S S increases with relative humidity (not shown)

data from Matsuki et al. (2010), 10.5194/acp-10-1057-2010 28 Mixing of dust with sulfate at Praia, Cape Verde

iron-rich grains silicate + Particles are sulfate shown before and after extensive electron bombardment

before 200 nm after

silicate + sulfate + ?

loss of volatile material

Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 29 Mixing of dust and organics observed in Senegal

SE image Si map C map particle and mesh support particle only particle (and support) visible

. Internal mixing of dust and carbonaceous matter is observed in a region with coexisting dust and biomass burning aerosol . Relative abundance of internally mixed particles versus pure ones is highly variable (5 to 50 % in this case study) . Carbonaceous matter is distributed homogeneously around the particle  most probably organic coating . Is coating reversible or not (high adsorption efficiency of clays for organics)?

Deboudt et al. 2010, doi: 10.1029/2010JD013921 30 Transport/Deposition stage – particular importance of iron and phosphorus DEPOSITION TRANSPORT EMISSION

terrestrial solar radiative impact wet processing radiative impact photochemistry cloud impact indirect radiative impact anthropogeneous/ biomass burning aerosol admixture

heterogeneous chemistry

dry removal sea-salt interaction wet removal

terrestrial geological basement ecosystem impact type of weathering surface transport human/plant/animal marine ecosystem impact health issues chemical processing emission meteorology 31

Iron and phosphorus as nutrients

. As of today, Fe and P seem to be the most important nutrients, but knowledge about dust Ca, Mg, K supply is sparse . For soluble Fe fraction (i. e. bio-availability), a range between 0.01 and 80 % is reported(1,2) . Fe solubility depends on source composition, atmospheric processing, already existing Fe concentrations, dust concentration, biological influences(2,3) . Fe solubility increases with decreasing pH especially for low pH values(4) . Fe solubility was found to increase with atmospheric processing intensity(5) . Most soluble Fe comes from clay minerals, not from Fe oxides(6) . Fe-rich nanoparticles can form after acidic solution of soil(7), which may be directly or at least at higher rates bio-available(2) . High aerosol Fe content in general does not mean similarly high bio-available Fe(5,6) . Total P content in Saharan dust between 0.04 and 1.7% (mostly below 1%)(8,9,10) . Acidification of aerosol (e. g. anthropogenic gas emissions) makes P more bio-available(11) . High relative humidity may be counterproductive in increasing P-availability due to more neutral pH(11) . P is assumed to be present as apatite, but also other phases are very probable

1Mahowald et al. 2005, doi: 10.1029/2004GB002402 2Baker et al. 2010, doi: 10.1016/j.marchem.2008.09.003 3Shi et al. 2011, doi: 10.1029/2010GB003837 4Cwiertny et al 2008, doi: 10.1029/2007JD009332 5Chen et al. 2004, doi: 10.1029/2003JD003958 6Journet et al. 2008, doi: 10.1029/2007GL031589 7Shi et al. 2009, doi: 10.1021/es901294g 8Guerzoni et al. 2005, doi: 10.1007/b107149 9Guieu et al. 2002, doi: 10.1029/2001JD000582 10Singer et al. 2003, doi: 10.1006/jare.2002.1023 11Nenes et al 2011, doi: 10.5194/acp-11-6265-2011 32

Phosphorus in Saharan dust

P 0° P Position of P on single dust particles

90° mirr.

in part from Scheuvens et al. 2011, doi: 10.1111/j.1600-0889.2011.00554.x 33 Deposition stage – variability and future needs DEPOSITION TRANSPORT EMISSION

terrestrial solar radiative impact wet processing radiative impact photochemistry cloud impact indirect radiative impact anthropogeneous/ biomass burning aerosol admixture

heterogeneous chemistry

dry removal sea-salt interaction wet removal

terrestrial geological basement ecosystem impact type of weathering surface transport human/plant/animal marine ecosystem impact health issues chemical processing emission meteorology 34 Deposition stage – phenomenology and dust composition

Type of deposition . In Florida, most dust is deposited by wet deposition(1) . At Bermuda, most dust is deposited by dry deposition(2) . Mixing with sea-salt could increase particle size and promote (dry) deposition(2,3,4,5)

Variability . High temporal variation in mass: 30 to 90% of annual dust deposition occurs on 5% of the days, particularly at the edge of the Saharan dust plume(6,7) . Low temporal variation in composition: two years of measurement at Florida and Barbados(8) and low spatial variation in composition over Florida(1) . Temporal variability (Fe/Ca) at Puerto Rico still in the same range as over the Ocean(9,10)

General . Dust arrives internally mixed with different species in America(5,11) . Information on speciation of major nutrients and their availability is sparse(12)

1Prospero et al 2010, doi: 10.1029/2009JD012773 2Tian et al. 2008, doi: 10.1029/2007GC001868 3Zhang et al. 2006, doi: 10.1016/j.atmosenv.2005.10.037 4Deboudt et al. 2010, doi: 10.1029/2010JD013921 5Worobiec et al. 2007, doi: 10.1016/j.atmosenv.2007.07.056 6Mahowald et al. 2009, doi: 10.1146/annurev.marine.010908.163727 7Bonnet et al. 2006, doi: 10.1029/2005JC003213 8Trapp et al. 2008, doi: 10.1016/j.marchem.2008.10.004 9Reid et al. 2003, doi: 10.1029/2002JD002935 10Kandler et al. 2011, doi: 10.1111/j.1600-0889.2011.00550.x 11Krejci et al. 2005, doi: 10.5194/acp-5-3331-2005 12Okin et al. 2011, doi: 10.1029/2010GB003858 35 Composition of aerosol deposited in the northern Amazonian basin

particle diameter 0.5 to 2 µm Organic 100%

90% Biogenic

80% Aged sea salt + soil dust 70%

60%

50% Aged sea salt + organic

40%

30%

20%

10% Soil dust + organic 0% 25 Mar 26 Mar 27 Mar 29 Mar 30 Mar 1 Apr 2 Apr 6 Apr 7 Apr 8 Apr 10 Apr 11 Apr 12 Apr 1998 soil dust refers to non-local (Saharan) dust, Sea salt + organic determined by trajectory analysis

Worobiec et al. 2007, doi: 10.1016/j.atmosenv.2007.07.056 36

What is missing?

General effects . Emission: We have seen considerable amounts of sulfate already on dust in Africa – what is the degree of processing at the soil surface, and how much is it processed during transport? . Emission/Transport: Regarding the Fe solubility and bio-availability, we need to know more on their dependence on source mineralogy, on in-soil and in-air processing (which gaseous or particular species yield what effect?), and how these impact on ecosystems with shorter or longer retention time . Emission/Transport: Similar questions arise for phosphorus availability, but the level of knowledge is even lower than for iron . Transport: For cloud impact and also the dust processing, we need to know more about the particle mixing state – preferrably spatially (3D) and particle-size-resolved . Transport/Deposition: We know that dust has an non-linear impact on cloud droplet and ice nucleation and, thus, precipitation, but which are the dominating effects? – Possibly by a combination of an intensive field campaign and subsequent quantification (monitoring) . Transport/Deposition: Besides Fe and P, also Ca, Mg and K from African dust are termed as potential nutrient for terrestrial ecosystems – what is their impact in relation to other sources? How changes their bioavailability by processing dependent of different atmospheric acids? . Transport/Deposition: We know that organic coatings exist on dust – do they derive from biomass burning (only), from marine processes, can they be acquired just before deposition? What do they do to cloud impact (CCN, IN properties), and can organic acids promote bioavailability in time? . Deposition: Does internal mixing increase deposition flux? Does particle shape have an impact?

What is missing? (continued)

Spatial and temporal variation . Dust sources have small-scale compositional variation and are not time-invariant  Is it feasible to create a adequately-resolved “cadaster”? What can we generalize?  What is the influence of this variation on the receptors (clouds, ecosystems)? . Several intensive field experiments with different foci yielded information on dust composition, but the longest ones lasted one season  How should they be rated, given a considerable variability on inter-seasonal scale (NAO) and the strong event-like occurrence of large dust loads? . Many measurements show high daily variation, measures for variability change as single days are excluded  measurements on too short time scales to capture variability  Need of long-term monitoring (in terms of composition, single particle measurements would be useful) With respect to complexity of dust interactions  “Supersite” monitoring – which locations can be set-up upgraded? . We know that dust plumes can be sharp-edged on inhomogeneous (particularly on vertical axis, but also on horizontal one), but we can only monitor at single spots continuously  Need of network-like observations (aircraft- and ship-based monitoring, e. g. CARIBIC package)